I am working on a project for my thesis, and it requires implementation of the structure from motion technique to create point cloud models of . However, is it possible to define the point cloud model numerically?
I think SLAM and PTAM algorithms are what you need. Thanks to them you'll be able to generate point cloud and reconstruct seen environment by moving your camera.